is climate change driving yellow-cedar decline on haida … · 2018. 3. 13. · haida gwaii is...

0 200 400 600 800 1000

0

20

40

O N D J F M A M J J A S O N D J F M A M J J A S

O N D J F M A M J J A S O N D J F M A M J J A S

1946-1976 1977-2014

Tave

PPT

Live Chronology (detrended with 100 yr spline)

Understory Overstory

0

100

200

300

200 400 600 800

0.0

0.4

0.8

Live Declining Dead

0

20

40

DENSITY (trees/ha)• Proportion of declining

and dead trees does not increase with density

• No evidence of density-dependent mortality

DBH (cm)• Mean DBH of live,

declining and dead trees were similar

• Declining and dead trees were not the smallest

• No evidence of self-thinning

IS CLIMATE CHANGE DRIVING YELLOW-CEDAR DECLINE ON HAIDA GWAII?Vanessa Comeau and Lori Daniels – The Tree-Ring Lab at UBC, Faculty of Forestry, University of British Columbia, Vancouver

BACKGROUND• Yellow-cedar decline is widespread along coastal BC and

Alaska and recently became apparent on Haida Gwaii• The driving factor of this decline is thought to be climatic

warming, which has reduced snowpack and exposed fine roots to freezing damage

• This proposed mechanism may not adequately explain the decline in Haida Gwaii due to the temperate climate and ephemeral snowpack

• This research uses dendrochronology to investigate climate as a driver of decline

• Determining the cause of this decline is the first step to resilience

QUESTIONS• In targeted forests exhibiting decline, what proportion of

yellow-cedar are live (healthy), declining or dead?

• Which yellow-cedar are declining and why?

• Does growth near the outer rings of live, declining and dead yellow-cedar differ?

• How did inter-annual and multi-decadal climate variation in the 20th century influence radial growth of yellow-cedar?

• Can this decline be mitigated through management?

RESEARCH APPROACH• Targeted areas exhibiting yellow-cedar decline

on Graham Island, Haida Gwaii• Census of all yellow-cedar trees (DBH ≥10cm) in

100 x 20m transects at 15 sites (1016 trees)• Increment cores taken from 15 healthy live and

15 declining/dead yellow-cedars at each site (450 trees)

MOST YELLOW-CEDAR DEAD OR DECLINING

DECLINE ACROSS SIZE AND AGE

DEATH AND CESSATION OF RADIAL GROWTH

SUMMARY• At targeted sites, high proportions of declining and dead trees• Trees of all ages and sizes are affected, with greater impacts on

overstory trees, but not due to competition and self-thinning • Trees of all status classes show suppression near the outer rings• Warm temperatures facilitated increasing growth in the 20th century• Decline in growth associated with low winter precipitation (snow) and

warm temperatures, is consistent with the drivers of decline in Alaska • Improved understanding of snow distribution and persistence in

Haida Gwaii is needed to guide management

ACKNOWLEDGEMENTSThank you for funding from Taan Forest, BC FLNRO and NSERC-Engage, collaboration with the Council of the Haida Nation and R. Chavardès (Tree Ring Lab at UBC), and research assistants: G. Knochenmus, I. Jarvis, J. Liu, A. Weixelman, J. Coutu, J. Heredia, D. Hornsberger, S. Bronson and D. Fluharty for assistance in the field and lab. VC received scholarships from NSERC-CGS and the UBC Faculty of Forestry.

• 1016 trees sampled at 15 targeted sites exhibiting yellow-cedar decline

• 74% of yellow-cedar were declining or dead • 59 to 86% of trees were declining or dead per site• 14 to 41% of trees were live, healthy per site

DECLINE NOT DUE TO COMPETITION

HEIGHT CLASS• More dead overstory trees

and fewer live and declining understory trees than expected by chance

• Opposite to patterns expected due to competition and self-thinning

AGE ESTIMATES• Age range: 29 to 908 years• Live, declining and dead trees present

across all ages • Declining and dead trees were not oldest

(reaching lifespan) or youngest (self-thinning)

DIAMETER AT BREAST HEIGHT• DBH range: 10 to 140 cm• Live, declining and dead trees present

across all size classes • Similar distributions of size for live,

declining and dead trees

• If winter precipitation is partly snow, relationships in the winter are consistent

with the snowpack hypothesis from Alaska

CLIMATE-GROWTH SHIFTS THROUGH TIME

Dead46.5%

Live26.1%

Declining27.5%

1946-1976Cool-dry-more snow

1977-2014Warm-wet-less snow

• Spring and summer correlations stronger from 1977-2014 in the warm phase

• Wet springs limit growth

• Precipitation in fall and winter (snow?) has more influence than in the growing season

Subset of 357 trees (live:208, declining:80, dead:69)

Freq

ue

ncy

(tr

ees)

Total Precipitation

Mean Temperature

p=0.16, 0.11, 0.65 p<0.001

Mea

n D

BH

(cm

)

Pro

po

rtio

n (

tree

s)

Co

rrel

atio

n c

oef

fici

ents

Pro

po

rtio

n (

tree

s)

SUPPRESSION AT OUTER RINGS• Decline is visible in radial growth rings before death in many trees• 13% of live and 46% of declining trees had suppressed outer rings• For trees which had died 24% had suppressed outer rings

Freq

ue

ncy

(tr

ees)

Age (years)n=444

20 40 60 80 100 120

0

20

40

60

80

100

120DBH (cm)n=1016

p=0.03

Freq

ue

ncy

(tr

ees)

1920 1940 1960 1980 2000

0

2

4

6

1985 1995 2005 2015

0

2

4

6 OUTER RING DATES• 191 live and 73 declining trees formed rings

the year of sampling• 59 living trees (live or declining) stopped

forming rings after 1918, 70% from 2000 to 2014

• 69 trees died after 1852, 42% from 2000 to 2015

Warm/wet Nov, wet Dec limit growth

Dry Nov/Dec, warm Jan limit growth

Warm spring and summer facilitates growth

marks 50yrs